Infant mortality among Arab-Americans: findings from the Arab-American birth outcomes study.

Arab-Americans (AAs) have lower risk of preterm birth relative to Non-Arab Whites. This has been attributed to lower likelihood of birth out of wedlock, maternal tobacco use during pregnancy, and foreign maternal birthplace among AAs. We were interested in understanding the roles of these and other demographic factors in the etiology of infant mortality among this group. Using data about all live, singleton births between 1989 and 2005 in the state with the highest proportion of AAs in the US, we calculated infant mortality (death prior to 1 year of life) for AAs and Non-Arab Whites. To clarify the etiology of potential differences in infant mortality, we also assessed infant mortality sub-categories, including neonatal mortality (death prior to 28 days of life) and post-neonatal mortality (death between 28 and 365 days of life). We fit trivariable and multivariable logistic regression models adjusted for explanatory covariates to assess each covariate's contributions to the relation between ethnicity and infant mortality. AAs had a lower infant mortality rate (4.7 per 1,000 live births) than non-Arab Whites (5.6 per 1,000 live births), overall (odds ratio = 0.84, 95% confidence interval: 0.74-0.96). In trivariable models, adjusting for marital status, maternal tobacco consumption during pregnancy, and maternal birthplace each separately attenuated the bivariate ethnicity-mortality relation to non-significance. Our findings suggest that lower risk of infant mortality among AAs relative to non-Arab Whites may be explained by differences in demographic characteristics and parental behavioral practices between them.